Riskified Ltd. (RSKD) Earnings Call Transcript & Summary

Okay, last but most certainly not least, to wrap up the conference, we have Eido Gal here, CEO of Riskified. So great to have you here. Thanks for coming. Appreciate it.

Thanks for inviting us.

Okay. Talk a little bit about -- again, a little bit late for this given that folks are probably quite familiar with the model, but nevertheless, give us a little overview of the company for those who are not familiar.

Sure. So for people who are not familiar, right, if I steal kind of Ramsey's credit card, while we're speaking right now, order a bunch of stuff online. He sees it on his statement, asks around at home did anyone order all these iPhones from Best Buy and everyone says no. He calls the bank, says, "hey, it wasn't me." The bank refunds, right? The U.S. consumers are protected with a credit card purchase. But the money gets taken out of the merchants account, right? So merchants are liable for accepting card-not-present transactions that have been from stolen financials, okay? And what that creates is the situation where every e-commerce merchant in the world, basically, also the significant enterprise funds are screening incoming transactions to decide if they're fraudulent or not, okay? And that creates a lot of inefficiencies. They accept a lot of bad transactions that result in loss. They turn away a lot of good customers, which is very frustrating and leads to missed bottom line. And then they also have a large population of transactions that sit in manual review and they [ can ] offer like same hour shipping and omnichannel flows because of the risk concerns. So that's the problem that we're solving. And when we started the company about a decade ago, we said, instead of all these individual merchants building and managing their own risk team and processing setup, we think we can do it better than they can, right? We have a better network effect, better machine learning, more resources towards solving that problem. And in order to really enable us to be an end-to-end solution, we need to offer them 2 things. Number one, is a chargeback guarantee, okay? So if we accept the transaction, approve a transaction and it turns out that we were wrong, we need to pay the merchant back. And the second thing we need to offer is an approval rate guarantee, right, so that the merchant knows that we'll approve more transactions than he has historically. And those 2 things enable the merchant to just kind of say, hey, I don't need to worry about this process anymore. Riskified is going to manage it for me.

I see, I see. And you guys have a pretty differentiated view of the e-commerce environment. There's -- it's been really interesting to watch e-commerce sort of macro e-commerce trends. I feel like there's this sort of normal cyclical cycle, but there's also kind of an e-commerce cycle, which is a little bit out of sync just because of maybe the pandemic and tough [ grow overs ] and this kind of odd goods versus services, pendulum that's been swinging back and forth. What's your view right now of the e-commerce environment just in terms of those types of forces and whether we're seeing some normalization coming?

You're right because the pandemic happened and then home goods and stay-at-home categories took off. Anything related to tickets, travel services went down. Reopenings are happening, right? And then you're seeing this normalization when tickets and travel, just everyone wants to make up for all that lost time, lost experiences. And when we think about some of the other COVID beneficiaries, I think everyone is saying, hey, we're still -- they're still digesting the growth that they went through during COVID. So what we're seeing and continue to see in this Q1 that we just released is that some of the categories like home and general or retail while they're not declining as much, still seeing some year-over-year declines, okay? Tickets and travel continues to be very strong. And now we have the overlay of an overall weakening economy and how is that influencing. And I think in there, we saw that there was a general slowdown in luxury spend as well.

And I guess as a follow-on to that, and you called out travel, I mean, what are you seeing travel specifically? I mean that's been to your point, just this huge boost coming out of the pandemic, people saying I don't want to get out of my home office and get back on the road, and also, obviously, from a consumer perspective as well, what are the trends like there? Are you still seeing some incremental rift? Or is it things kind of starting to crest?

We're still seeing a lot of sustained growth. So when we think about Q1, I think our growth was almost 100% within tickets and travel now. This isn't just organic same-store sales growth. We've also added a lot of merchants in the tickets and travel industry so that's also helping us drive some of that demand. We do anticipate some normalization towards the end of the year, consistent with what these kind of public companies have been saying and disclosing themselves. But we still feel it's very, very strong. It's also important to mention, the tickets is events, it's Taylor Swift, it's NBA. So there's a lot of strength there, and I think that strength is going to continue. People have shifted more budget towards those areas long term. Travel, I think people are making up for a lot of loss travel opportunities. I don't know at what rate it would sustain into the future, but we're still seeing a lot of strength there.

As you called out when you're describing your business, the chargeback guarantee is a central part of your value proposition. How do you -- you sort of manage risk around that product. How do you underwrite risk at the right kind of equilibrium of increasing approvals while still preventing the fraudulent transactions? What is that process like internally?

I mean the way to think about it is, look, we believe that we've created the most accurate decisioning platform to understand if a transaction would result in a chargeback or not. And we're not really underwriting this unknown probability that something like a hurricane would happen, right? We're saying we've identified that this is like a fraudster that's stolen this card or not. And we just have, through our network effect, to our better resources. And now, we just have more accuracy on that. Now the chargeback guarantee, it's more of a method so that the merchant can trust us right, to give us ownership and control over that part of the business. Now what we're actually doing, we're doing millions, hundreds of millions of smaller transactions on an ongoing basis, right? We're not guaranteeing a $90 million swift transfer between geographies or anything like that. And we continuously have an internal expectation of how many chargebacks we should be receiving within what time frame. And then you can know that as soon as a few days. And what we always see is are things tracking to what we anticipate or maybe for a specific merchant, for a specific model in a specific use case, we see that maybe we're a bit off, and then we need to course correct. But the time frame where chargebacks happen, okay, and chargebacks come back in within a few days. They fully mature within a few months, but they start coming back in within a few days. The fact that those transactions are for lower average dollar amounts, usually a few hundred dollars, gives us a lot of safety and large numbers and the ability and timing to understand if anything is off and course correct based on that.

Interesting. So you really do have quite a bit of visibility. It's not something that emerges really abruptly. It's something where you see the signals pretty early and it gives you time to basically...

Yes, I think that's a great way to think about it.

Interesting. There's something I've always been meaning to ask you. Is there any kind of correlation between chargeback levels and economic stress? Is there -- is it more difficult to predict chargebacks as -- in different parts of the economic cycle? Or is that not the case?

I think from -- it's pretty consistent in good times and bad times, right? So there's always been online fraud and it's increasing in sophistication. I've never been able to really pinpoint and say, hey, in a tougher macro, we're seeing an increase of fraud, and we've looked into that at various points. This isn't like defaults, right? You would expect defaults and interest rates, they have an unbelievably clear impact, right? But when you think about just the ability to create fraud, it's somewhat consistent.

That makes a lot of sense. Talk to us about the competitive environment. Who are you competing with? And maybe don't just limit the thinking to other companies who might be offering something similar -- the fintechs kind of offering something similar. What is the broader view of competition in-house solutions, that type of a lens?

So the vast majority of merchants, they have an in-house team, that in-house team uses a number of third-party providers, and they could be providers from like a data enrichment provider. It's a provider that gives them the tools to do rule building, machine learning building, case management, but these are tools used by the internal team, right? And you really -- Riskified says, hey, instead of having a team using these various tools, we think we can perform better and replace that entire construct. Now these tools that the teams are using, 90-plus percent of the time, they are legacy tools, older tools. Sometimes they could be some of the newer generation players, right? There's kind of probably 2 or 3 of those out today. I think that altogether, the market share for the newer players is still sub-10%, if I had to size it. And outside of dedicated risk solutions, and we're a dedicated risk solution, that's what we do. You also have payment gateway solutions in the enterprise space, where we play, they're much less prominent, right? And there's a few reasons for that. One of them being that enterprises tend to run multiple payment gateways and route transactions and have fail safe and wherever they have better margins and unit economics there. Second being that most of the payment gateways have much limited data integration points because they want everyone to get up and running in no time and they don't want to create a very complex integration cycle, whereas we tend to go a bit deeper, which helps our modeling and performance overall.

And when you do sign a new customer, how does that typically -- how does the sort of life cycle of that relationship work? Do they say to you, hey, we want you to take our full book? Do they give you parts of the flow that are more problematic where they need some help? Or how does it start versus where you kind of get to that customer, what's that journey look like?

So let's start with us, hey, we're talking -- and as you're managing the payments and risk team for a $20 billion OTA, right? And we start to map out the different flows and segments, you have a few different product lines, you have [ accommodations ], you have flights, you have [indiscernible], you're operating in APAC and the U.S. You have some declines, you have some transactions that are routed to manual or really have a bunch of stuff, right? And we take a CSD for you with all that data and we run a different analysis and we come to you and say, hey, you know what, we think about your APAC traffic. Actually, you're paying 50 basis points for a 90% approval rate, but we could do that for 35 basis points, okay? Or if you want, for 50 basis points, we can provide you a 95% approval rate, right? So we find segments where we can overperform the most on an existing merchant setup. And in that sense, we're not asking you to replace everything you have. We're augmenting you in one area where we think we can do better. So we do an integration, okay? We start to build this trusted relationship. You start to see that we can perform. Now we're only -- in this example, we're only doing APAC for you, but we see and capture all of the rest of the volume, okay? So we can also analyze what's going on in the rest of your business by now. And then maybe a few months down the line, we're going to come and we're going to say, hey, you know what, the rest of the business or these other parts that you're not giving to us today, we can actually do a much better job and here's the economics, and here's the business case, and it's very, very ROI-driven. And that's the standard land-and-expand process that we have.

And how does kind of new products and cross-sell sort of fit into that kind of schema? Maybe talk about -- describe for us some of the ancillary products that you have and how that kind of cross-sell process works? How does that fit into that roadmap?

Sure. So I'll talk about the one that I'm most excited about, we call it policy. And Policy Protect has a few different use cases. Maybe the one to highlight is around Refund Request, right, Refund Request and Return Requests. So what we found is that we can use the same ML platform network effects, the data that we have, not just to predict if a transaction would result in a charge back, but also to predict if when a consumer is coming and saying to the merchant, hey, I never received my package, I received the wrong size. I received the wrong whatever, it was damaged. I want a refund. Is this a legitimate claim, is this a legitimate consumer? Or is it just someone who says, you know what, even easier than to steal Ramsey's credit card is just to call the merchant and say, I never received my package, right? It's hard to prove -- [ disprove ] anything, 9 times out of 10, a merchant would honor that refund request, that return request. And when we run the analysis, we find that there's a ton of fraudulent activity and behavior there. So 2 recent examples with merchants that went live, we were able to increase their reject rate from less than 2% to 15%, while not increasing the insult rate. And the insult rate is the amount of good customers that you're turning away and that's measured by either callbacks or actual chargebacks after they decline them. So I think there's massive value in those areas as well. And that's the one that I'm kind of more excited about. The way we go to market with it is a combination of talking to -- when we talk to newer clients, we talk to them about the platform more so than just the fraudulent chargebacks. And when we think about our existing clients, it's definitely a cross-sell opportunity.

And what's the economic model on a product like that is -- just very simply, is it functioned somewhat similarly to the -- to your core product? Or how does it...

It's slightly different. So the core product is more -- it's usage-based on the GMV flowing through, and there's the guarantee component. Here, it's just a much simpler SaaS, monthly subscription, obviously, based on volume and size of merchant. And the margins themselves are -- also they don't have the guarantee components, so they're more traditional SaaS like market.

I see. Okay. And if you fast forward a few years in terms of the product -- overlaying a product roadmap to your kind of growth algorithm or your revenue mix, I guess, I should say. How important do you think some of these ancillary products are going to become? Is this always going to be the case where your core is the chargeback type kind of protection services and then you're sort of hanging some stuff off the side? Or does any of this have the potential to really open up a big aperture in terms of being a really primary growth driver over time?

It depends on how many years forward and how successful we're not selling it, right? But I think when I think about the business, I don't think about chargebacks, right? I think about we have a very deep integration into the online system, into the off-line system, [indiscernible] system, the customer support system because we need all that to make smart fraud decisions. And internally, we have a great machine learning platform capabilities that are really focused and dedicated towards e-commerce, which is unique. And at any point where an enterprise e-commerce needs to make a decision, right? Do I honor this request with this result and charge back, what's the LTV of this customer whatnot? There's a lot of decisions that they need to make that I think we're incredibly well situated to answer and solve for them. right? So when I think about the fact that we've had kind of 99% gross retention rate for a few years now. Our customers love us. They stay with us for a long time. We're always thinking what other problems or questions can we help solve for them using our existing technology and stack. So I think this approach will lead us over time to situation where the revenue that is associated with these other products outside of our chargeback guarantee is significant. What is that time frame? How long does that scale? That's -- we need to work on that.

Fair enough. I meant to ask you this a little bit earlier. What if merchant [ acquirers ]?

I think they'll prefer someone like us. I don't think they view us very competitive. I think that they understand that our value offering is distinct enough from their real core value offering, not just what they maybe talk about doing. I think a lot of them like to bring us in when they need someone to help their merchants that are having issues.

So talk about that a little bit. So you have a kind of a partnership type relationship with some of these acquirers who will effectively pull you into the conversation when there's initiative.

When they know that there's an issue, for sure. Those are [ my ] comments.